We have continued our studies of chromatin structure in the neighborhood of expressed genes. The globin gene family in chicken erythroid cells serves as a model system in which it is possible to study the mechanisms associated with regulation of the individual members of the family during erythroid development. We have cloned the gene for the general erythroid-specific factor GATA-1 (Eryf1) of chicken. We find that the promoter of the gene contains binding sites for a number of known trans-acting factors, including a cluster of sites for GATA-1 itself. Transactivation studies show that these sites function in an autoregulatory mechanism. Other features of the promoter suggest mechanisms that may function early in development. We have continued studying the stage-specific expression of globin genes. We find that the pi-globin gene, an embryonic alpha-family gene, is controlled by a complex mechanism involving three factors that change their concentration at the embryonic-adult switch. We have continued studies of the effects of a beta-globin locus control region (LCR or DCR) on gene expression and chromatin structure. The LCR confers position independence of expression on nearby genes. We find that it is also capable of generating a hypersensitive domain in the absence of other elements, while elements like the beta-globin promoter are not, a distinction important in understanding how active chromatin is generated. In order to understand how the histones of chromatin function during transcription, we have also studied the effect of positive supercoiling on nucleosome core formation. We find that core particles formed on such DNA are perfectly normal, but spontaneously transfer to negatively supercoiled DNA when that is available. We have proposed a mechanism, based on these observations, for transcription through nucleosome-covered DNA.